Research map / Video / Video generation
Mask2IV: Interaction-Centric Video Generation via Mask Trajectories
Gen Li, Bo Zhao, Jianfei Yang, Laura Sevilla-Lara
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| concatenate mask latent with noise instead of ControlNet (ablation) | Mask2IV · HOI4D (GT mask trajectories) | FVD: 157.38 -> 130.07 | Improved |
| add object mask to the trajectory (ablation) | Mask2IV · HOI4D (GT mask trajectories) | FVD: 130.07 -> 115.14 | Improved |
| add random dilation/erosion perturbation of masks (ablation) | Mask2IV · HOI4D (GT mask trajectories) | FVD: 115.14 -> 108.80 | Improved |
| add contact weighting loss (ablation) | Mask2IV · HOI4D (GT mask trajectories) | FVD: 108.80 -> 104.61 | Improved |
| add contact weighting loss (ablation) | Mask2IV · HOI4D (GT mask trajectories) | LPIPS: 0.124 -> 0.126 | No clear effect |
| fine-tune DynamiCrafter on the curated data | DynamiCrafter · HOI4D | FVD: 554.48 -> 168.73 | Improved |
| two-stage mask trajectory control vs fine-tuned DynamiCrafter | Mask2IV · HOI4D | FVD: 168.73 -> 149.68 | Improved |
| two-stage mask trajectory control vs fine-tuned DynamiCrafter | Mask2IV · BridgeData V2 | FVD: 197.82 -> 155.73 | Improved |
| Mask2IV vs CosHand with same pseudo control | Mask2IV · BridgeData V2 | T2V-Sim: 0.220 vs 0.220 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DynamiCrafter — DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors (ECCV 2024) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Through-The-Mask: Mask-based Motion Trajectories for Image-to-Video Generation (CVPR 2025 · similarity 0.90)
- MIMIC: Mask-Injected Manipulation Video Generation with Interaction Control (ICLR 2026 · similarity 0.90)
- Learning Video Generation for Robotic Manipulation with Collaborative Trajectory Control (ICLR 2026 · similarity 0.89)
- Learning to Generate Rigid Body Interactions with Video Diffusion Models (ECCV 2026 · similarity 0.88)
- InterMask: 3D Human Interaction Generation via Collaborative Masked Modeling (ICLR 2025 · similarity 0.88)
- TokenMotion: Decoupled Motion Control via Token Disentanglement for Human-centric Video Generation (CVPR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
